基于视频检测系统的城市道路交通状态估计

发布时间：2018-06-16 14:21

本文选题：视频检测 + 行程时间　；参考：《山东大学》2015年硕士论文

【摘要】：随着城市交通的不断发展,智能交通管理系统在城市交通管理中发挥了越来越重要的作用,而其中城市道路卡口监控系统的广泛应用又推动了智能交通的快速发展。卡口系统具有目标车辆捕捉、车牌识别、断面车流统计、违法抓拍等功能。卡口系统已经成为智能交通管理系统中的重要组成部分,城市和城市交通管理者也越来越重视通过卡口系统进行数据的深度挖掘与利用。交通状态是交通信息服务系统的基础,城市道路实时交通信息是实现交通管理及控制、动态诱导、以及改善道路通行条件的基础,对城市规划、公交调度、市民出行等均具有重要参考价值。现有交通状态估计大都是基于数理统计或者模式识别等方法对交通状态进行划分,但这些划分方法都是主观的,相关研究缺乏主观与客观的结合。本文以道路实际行程时间信息为基础数据,结合驾驶员期望建立交通状态评价方法,实现了主客观的结合。整个论文包括以下几个部分。首先,卡口系统作为智能交通管理系统的重要部分,已经广泛的应用于我国大部分城市。通过卡口系统的车牌识别功能可以获得同一辆车通过不同检测点位的时间信息,由此对车牌号码进行匹配计算得到车辆的路段行程时间。由于检测设备自身的工作原理以及交通流的自主行为,造成采集的数据出现错误,主要表现为数据丢失和数据异常。本文利用卡口系统采集的文本信息创建数据信息库,针对不同情况下的数据缺失进行合理的修复,采用类似于t检定或者z检定的中位值偏差法对异常数据进行处理,经过处理后的数据计算得到车辆的路段行程时间样本集。准确的行程时间样本是交通状态估计的基本前提,交通状态的合理估计还需要建立完善的评价指标。基于行程时间的指标很容易被交通参与者理解,行程时间可以从不同的时间、空间维度上按不同要求描述交通状况；行程时间指标既能恰当的描述特定地点交通拥挤状况,也能描述整个道路交通拥挤状况。现有交通状态指标均假设不同道路等级的交通状态具有一致性,当单个路段或单个车辆的交通状态指标相同时则认为拥堵程度也相同,但是从管理者和驾驶员的角度讲,在不同等级道路的路段上,交通状态相同时管理者和驾驶员的感受是不相同的,例如在快速路和干道上速度同样降低20%,管理者和驾驶员认为在快速路更为拥堵,主要是因为出行者在快速路上期望的绝对行驶速度和相对行驶速度均比较高,所以不同路段或单个车辆计算交通状态时需要考虑驾驶员的感受。本文对通过对驾驶员进行期望速度调查,以路段行程时间为基础,结合驾驶员期望速度,制定了交通状态评价指标。最后,本文将将交通状态分为畅通、轻度畅通、轻度拥堵、中度拥堵、严重拥堵五个等级,并编程实现交通状态的显示。
[Abstract]:With the continuous development of urban traffic, intelligent traffic management system (its) has played an increasingly important role in urban traffic management, and the wide application of urban road bayonet monitoring system has promoted the rapid development of intelligent transportation. The bayonet system has the functions of target vehicle capture, license plate recognition, cross-section traffic flow statistics, illegal capture and so on. The bayonet system has become an important part of the intelligent traffic management system, and city traffic managers pay more and more attention to the depth mining and utilization of data through the bayonet system. The traffic state is the foundation of the traffic information service system, and the real-time traffic information of the city road is the basis of realizing the traffic management and control, the dynamic guidance, and the improvement of the road traffic conditions. The citizen travel and so on all has the important reference value. Most of the existing traffic state estimation methods are based on mathematical statistics or pattern recognition, but these methods are subjective and lack of the combination of subjective and objective. Based on the information of road travel time and the expectation of drivers, this paper establishes a method of traffic condition evaluation, and realizes the combination of subjective and objective. The whole paper includes the following parts. Firstly, as an important part of intelligent traffic management system, bayonet system has been widely used in most cities of our country. Through the license plate recognition function of the bayonet system, the time information of the same vehicle passing through different detection points can be obtained, and then the vehicle section travel time can be calculated by matching the license plate number. Because of the working principle of the detection equipment and the independent behavior of the traffic flow, there are errors in the collected data, which are mainly reflected in the data loss and data anomalies. In this paper, the text information collected by the bayonet system is used to create the data information database, and the data missing in different cases is repaired reasonably. The method of midpoint deviation similar to t test or z test is used to deal with the abnormal data. The processed data are calculated to get the sample set of the road travel time of the vehicle. Accurate travel time sample is the basic premise of traffic state estimation. The index based on travel time is easy to be understood by traffic participants. Travel time can be described according to different requirements in different time and space dimensions. It can also describe the whole road traffic congestion. The existing traffic state indicators all assume that the traffic state of different road grades is consistent. When the traffic condition index of a single road section or a single vehicle is the same, the traffic congestion degree is the same, but from the perspective of managers and drivers, On different levels of roads, managers and drivers feel differently when the traffic is in the same state. For example, on expressways and trunk roads, the speed is also reduced by 20 percent, and managers and drivers think that it is more congested on expressways. The main reason is that both the absolute speed and the relative speed expected by the traveler on the expressway are relatively high, so it is necessary to consider the driver's feeling when calculating the traffic state of different sections or individual vehicles. In this paper, through the investigation of the expected speed of the driver, based on the travel time of the road, combined with the expected speed of the driver, the evaluation index of the traffic condition is established. Finally, this paper will divide the traffic state into five grades: smooth, mild, moderate and severe congestion, and realize the display of traffic state by programming.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U491;TP274

【相似文献】